• Share
  • Email
  • Embed
  • Like
  • Save
  • Private Content
Cmos uma
 

Cmos uma

on

  • 1,043 views

 

Statistics

Views

Total Views
1,043
Views on SlideShare
1,043
Embed Views
0

Actions

Likes
1
Downloads
25
Comments
0

0 Embeds 0

No embeds

Accessibility

Categories

Upload Details

Uploaded via as Adobe PDF

Usage Rights

© All Rights Reserved

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Processing…
Post Comment
Edit your comment

    Cmos uma Cmos uma Document Transcript

    • Unit 2BASIC PHYSICAL DESIGN AN OVERVIEWThe VLSI design flow for any IC design is as follows1.Specification (problem definition)2.Schematic(gate level design) (equivalence check)3.Layout (equivalence check)4.FloorPlanning5.Routing,Placement6.On to SiliconThe preceding lectures have already given you the information of the different layers,their representation (colour,hatching)etc. When the devices are represented usingthese layers, we call it physical design. The design is carried out using the design tool,which requires to follow certain rules. Physical structure is required to study theimpact of moving from circuit to layout. When we draw the layout from theschematic, we are taking the first step towards the physical design.Physical design is an important step towards fabrication. Layout is representation of aschematic into layered diagram. This diagram reveals the different layers like ndiff,polysilicon etc that go into formation of the device.At every stage of the physical design simulations are carried out to verify whether thedesign is as per requirement. Soon after the layout design the DRC check is used toverify minimum dimensions and spacing of the layers. Once the layout is done, alayout versus schematic check carried out before proceeding further. There aredifferent tools available for drawing the layout and simulating it. The simplest way to begin a layout representation is to draw the stick diagram. But asthe complexity increases it is not possible to draw the stick diagrams. For beginners iteasy to draw the stick diagram and then proceed with the layout for the basic digitalgates . We will have a look at some of the things we should know before starting thelayout.In the schematic representation lines drawn between device terminals representinterconnections and any no planar situation can be handled by crossing over. But inlayout designs a little more concern about the physical interconnection of differentlayers. By simply drawing one layer above the other it not possible to makeinterconnections, because of the different characters of each layer. Contacts have to bemade whenever such interconnection is required. The power and the groundconnections are made using the metal and the common gate connection using thepolysilicon. The metal and the diffusion layers are connected using contacts. Thesubstrate contacts are made for same source and substrate voltage. which are notimplied in the schematic. These layouts are governed by DRC’s and have to be atleastof the minimum size depending on the technology used . The crossing over of layersis another aspect which is of concern and is addressed next.1.Poly crossing diffusion makes a transistor2.Metal of the same kind crossing causes a short.3.Poly crossing a metal causes no interaction unless a contact is made.
    • Different design tricks need to be used to avoid unknown creations. Like acombination of metal1 and metal2 can be used to avoid short. Usually metat2 is usedfor the global vdd and vss lines and metal1 for local connections.SCHEMATIC AND LAYOUT OF BASIC GATES1.CMOS INVERTERNOT GATE) SCHEMATICTOWARDS THE LAYOUT VIN VOUT Figure2: Stick diagram of inverterThe diagram shown here is the stick diagram for the CMOS inverter. It consists of aPmos and a Nmos connected to get the inverted output. When the input is low, Pmos(yellow)is on and pulls the output to vdd, hence it is called pull up device. When Vin=1,Nmos (green)is on it pulls Vout to Vss, hence Nmos is a pull down device. The redlines are the poly silicon lines connecting the gates and the blue lines are the metallines for VDD(up) and VSS (down).The layout of the cmos inverter is shown below.Layout also gives the minimum dimensions of different layers, along with the logicalconnections and main thing about layouts is that can be simulated and checked forerrors which cannot be done with only stick diagrams.
    • Figure 3: Layout of an inverterThe layout shown above is that of a CMOS inverter. It consists of a pdiff (yellowcolour) forming the pmos at the junction of the diffusion and the polysilicon(redcolour)shown hatched ndiff(green) forming the nmos(area hatched).The differentlayers drawn are checked for their dimensions using the DRC rule check of the toolused for drawing. Only after the DRC(design rule check) is passed the design canproceed further. Further the design undergoes Layout Vs Schematic checks andfinally the parasitic can be extracted. Figure 4:Schematic diagrams of nand and nor gateWe can seen that the nand gate consists of two pmos in parallel which forms the pullup logic and two nmos in series forming the pull down logic.It is the complementaryfor the nor gate. We get inverted logic from cmos structures. The series and parallelconnections are for getting the right logic output. The pull up and the pull downdevices must be placed to get high and low outsputs when required.
    • Figure 5: Stick diagram of nand gate Figure 6: Layout of a nand gateFigure 7:Stick diagram of nor gate
    • Figure 8: Layout of nor gateTRANSMISSION GATEFigure 9 :Symbol and schematic of transmission gateLayout considerations of transmission gate. It consist of drains and the sources of theP&N devices paralleled. Transmission gate can replace the pass transistors and hasthe advantage of giving both a good one and a good zero. Figure 10: Layout of transmissuion gate
    • Figure 11:TG with nmos switchesCMOS STANDARD CELL DESIGNGeometric regularity is very important to maintain some common electricalcharacteristics between the cells in the library. The common physical limitation is tofix the height and vary the width according to the required function. The Wp and Wnare fixed considering power dissipation, propagation delay, area and noise immunity.The best thing to do is to fix a required objective function and then fix Wn and Wp toobtain the required objectiveUsually in CMOS Wn is made equal to Wp . In the process of designing these gatestechniques may be employed to automatically generate the gates of common size.Later optimization can be carried out to achieve a specific feature. Gate array layoutand sea of gate layout are constructed using the above techniques. The gate arraysmay be customized by having routing channels in between array of gates. The gatearray and the sea of gates have some special layout considerations. The gate arraysuse fixed image of the under layers i.e the diffusion and poly are fixed and metal areprogrammable. The wiring layers are discretionary and providing the personalizationof the array. The rows of transistors are fixed and the routing channels are providedin between them. Hence the design issues involves size of transistors, connectivity ofpoly and the number of routing channels required.Sea of gates in this style continuous rows of n and p diffusion run across the masterchip and are arranged without regard to the routing channel. Finally the routing isdone across unused transistors saving space.GENERAL LAYOUT GUIDELINES1.The electrical gate design must be completed by checking the followinga.Right power and ground suppliesb.Noise at the gate inputc.Faulty connections and transistorsd.Improper ratiosc.Incorrect clocking and charge sharing2.VDD and the VSS lines run at the top and the bottom of the design
    • 3.Vertical poysilicon for each gate input4.Order polysilicon gate signals for maximal connection between transistors5.The connectivity requires to place nmos close to VSS and pmos close to VDD6.Connection to complete the logic must be made using poly,metal and even metal2The design must always proceeds towards optimization. Here optimization is attransistor level rather then gate level. Since the density of transistors is large ,wecould obtain smaller and faster layout by designing logic blocks of 1000 transistorsinstead of considering a single at a time and then putting them together. Densityimprovement can also be made by considering optimization of the other factors in thelayoutThe factors are1.Efficient routing space usage. They can be placed over the cells or even in multiplelayers.2.Source drain connections must be merged better.3.White (blank) spaces must be minimum4.The devices must be of optimum sizes.5.Transperent routing can be provided for cell to cell interconnection, this reducesglobal wiring problemsLAYOUT OPTIMIZATION FOR PERFORMANCE1.Vary the size of the transistor according to its position in series. The transistorclosest to the output is the smallest. The transistor nearest to the VSS line is thelargest. This helps in increasing the performance by 30 %. A three input nand gatewith the varying size is shown next.Figure 12 :Layout optimization with varying diffusion areas2. Less optimized gates could occur even in the case of parallel connectedtransistors.This is usually seen in parallel inverters, nor & nand.When drains are
    • connected in parallel ,we must try and reduce the number of drains in parallel iewherever possible we must try and connect drains in series at least at the output.Thisarrangement could reduce the capacitance at the output enabling good voltage levels.One example is as shown next.Figure 13 Layout of nor gate showing series and parallel drainsUNIT 3CMOS LOGIC STRUCTURESThe various application that require logic structures have different optimizations.Some of the circuit need fast response, some slow but very precise response, othersmay need large functionality in a small space and so on. The CMOS logic structurescan be implemented in alternate ways to get specific optimization. Theseoptimizations are specific because of the trade off between the n number of designparameters.CMOS COMPLEMENTARY LOGICWe have seen the logic CMOS structures of nand & nor. They were ratioed logic i.ethey have fixed ratio of sizes for the n and the p gates. It is possible to have ratiolesslogic by varying the ratio of sizes which is useful in gate arrays and sea of gates.Variable ratios allow us to vary the threshold and speed .If all the gates are of thesame size the circuit is likely to function more correctly. Apart from this the supplyvoltage can be increased to get better noise immunity. The increase in voltage mustbe done within a safety margin of the source –drain break down. Supply voltage canbe decreased for reduced power dissipation and also meet the constraints of the supplyvoltage. Some times even power down with low power dissipation is required. For allthese needs an on chip voltage regulator is required which may call for additionalspace requirement. A CMOS requires a nblock and a pblock for completion of thelogic. That is for a n input logic 2n gates are required. The variations to this circuitcan include the following techniques reduction of noise margins and reducing thefunction determining transistors to one polarity.BICMOS LogicThe CMOS logic structures have low output drive capability. If bipolar transistors areused at the output the capability can be enhanced. Bipolar transistors are currentcontrolled devices and produces larger output current then the CMOS transistors. Thiscombined logic is called BICMOS logic. We can have the bipolar transistors both forpull up and pull down or only for pull up as shown in the figures below. The figurenext shows a cmos nand gate with NPN transistors at both level.
    • The N1 & N2 supply current to the base of the NPN2 transistor when the out put ishigh and hence the it can pull it down with larger speed. When the output is low N3clamps the base current to NPN2, P1 & P2 supply the base current to NPN1.Figure 14 Nand with two NPN driversThis design shown previously is basically used for speed enhancing in highlyautomated designs like gate arrays. Since the area occupied by the Bipolar transistorsis more and if the aim in the design is to match the pull up and pull down speeds thenwe can have a transistor only in the pull up circuit because p devices are slower asshown in the figure next. The usage of Bicmos must be done only after a trade off ismade between the cost, performance etc.Figure 14 Nand with one NPN in pull up
    • PSEUDO NMOS LOGICThis logic structure consists of the pull up circuit being replaced by a single pull uppmos whose gate is permanently grounded. This actually means that pmos is all thetime on and that now for a n input logic we have only n+1 gates. This technology isequivalent to the depletion mode type and preceded the CMOS technology and hencethe name pseudo. The two sections of the device are now called as load and driver.The ßn/ßp (ßdriver/ßload) has to be selected such that sufficient gain is achieved toget consistent pull up and pull down levels. This involes having ratioed transistorsizes so that correct operation is obtained. However if minimum size drivers are beingused then the gain of the load has to be reduced to get adequate noise margin.There are certain drawbacks of the design which is highlighted next1.The gate capacitance of CMOS logic is two unit gate but for psuedo logic it is only one gate unit.2.Since number of transistors per input is reduced area is reduced drastically.The disadvantage is that since the pmos is always on, static power dissipation occurswhenever the nmos is on. Hence the conclusion is that in order to use psuedo logic atrade off between size & load or power dissipation has to be made.Figure 15 Pseudo NmosOTHER VARIATIONS OF PSEUDO NMOS1.Multi drain logicOner way of implementing pseudo nmos is to use multidrain logic. It represents amerged transistor kind of implementation. The gates are combined in an open drainmanner, which is useful in some automated circuits. Figure 16
    • GANGED LOGICThe inputs are separately connected but the output is connected to a commonterminal. The logic depends on the pull up and pull down ratio. If pmos is able to overcome nmos it behaves as nandelse nor.DYNAMIC CMOS LOGICFigure 17 Dynamic cmos logicThis logic looks into enhancing the speed of the pull up device by precharging theoutput node to vdd. Hence we need to split the working of the device into prechargeand evaluate stage for which we need a clock. Hence it is called as dynamic logic. Theoutput node is precharged to vdd by the pmos and is discharged conditionally throughthe nmos. Alternatively you can also have a p block and precharge the n transistor tovss. When the clock is low the precharge phase occurs. The path to vss is closed bythe nmos ie the ground switch . The pull up time is improved because of the activepmos which is already precharged. But the pull down time increases because of theground switch .There are a few problems associated with the design, like
    • 1.Inputs have to change during the precharge stage and must be stable during theevaluate. If this condition cannot occur then charge redistribution corrupts the outputnode.2.A simple single dynamic logic cannot be cascaded. During the evaluate phase thefirst gate will conditionally discharge but by the time the second gate evaluates, thereis going to be a finite delay. By then the first gate may precharge.CLOCKED CMOS LOGIC (C2MOS)Figure 18 C2mos logicCMOS DOMINO LOGICThe disadvantage associated with the dynamic CMOS is over come in this logic. Inthis we are able to cascade logic blocks with the help of a single clock. The prechargeand the evaluate phases retained as they were. The change required is to add a bufferat the end of each stage.This logic works in the following manner. When the clk=0,ie during the prechargestage the output of the dynamic logic is high and the output of the buffer is low. Sincethe subsequent stages are fed from the buffer they are all off in the precharge stage.When the gate is evaluated in the next phase, the output conditionally goes low andthe output of the buffer goes high. The the subsequent gates make a transition fromhigh to low.Figure 19:Cmos domino logic
    • Hence in one clock cycle the cascaded logic makes only one transition from 1 to 0 andbuffer makes a transition from 0 to 1.In effect we can say that the cascaded logic fallslike a line of dominos, and hence the name. The advantage is that any number of logicblocks can be cascaded provided the sequence can be evaluated in a single clockcycle. Single clock can be used to precharge and evaluate all the logic in a block. Thelimitation is that each stage must be buffered and only non- inverted structures arepossible.A further fine tuning to the domino logic can also be done. Cascaded logic can nowconsist of alternate p and n blocks and avoid the domino buffer. When clk=0,ie duringthe precharge stage, the first stage (with n logic) is precharged high and the second a plogic is precharged low and the third stage is high. Since the second stage is low, the ntransistor is off. Hence domino connections can be made.The advantages are we can use smaller gates, achieve higher speed and get a smoothoperation. Care must be taken to ensuredesign is correct.NP DOMINO LOGIC (ZIPPER CMOS)Figure 20: NP domino logicCASCADED VOLTAGE SWITCH LOGICIt is a differential kind of logic giving both true and complementary signal outputs.The switch logic is used to connect a combinational logic block to a high or a lowoutput. There are static and dynamic variants .The dynamic variants use a clock. Thestatic version (all the figures to shown next) is slower because the pull up deviceshave to over come the pull down devices. Hence the clocked versions with a latchingsense amplifier came up. These switch logic are called sample set differential logic
    • STATIC CVSLFigure 21 Static CVSLDYNAMIC CVSLFigure 22:Dynamic CVSL
    • DYNAMIC SSDL CVSLFigure 23:Dynamic SSDLCVSLPASS TRANSISTOR LOGICSwitches and switch logic can be formed from simple n or p transistors and from thecomplementary switch ie the transmission gate. The complex transmission gate cameinto picture because of the undesirable threshold effects of the simple pass transistors.Transmission gate gives good non degraded logic levels. But this good package cameat the cost of larger area and complementary signals required to drive the gates
    • Figure 24: Some properties of pass transistorCMOS Technology Logic Circuit StructuresMany different logic circuits utilizing CMOS technology have been invented and usedin various applications. These can be divided into three types or families of circuits:1.Complementary LogicStandard CMOSClocked CMOS (C2MOS)BICMOS (CMOS logic with Bipolar driver)2.Ratio Circuit LogicPseudo-NMOSSaturated NMOS LoadSaturated PMOS LoadDepletion NMOS Load (E/D)Source Follower Pull-up Logic (SFPL)
    • 3.Dynamic Logic:CMOS Domino LogicNP Domino Logic (also called Zipper CMOS)NOR A LogicCascade voltage Switch Logic (CVSL)Sample-Set Differential Logic (SSDL)Pass-Transistor LogicThe large number of implementations shown so far may lead to a confusion as to whatto use where. Here are some inputs1.Complementary CMOSThe best option,because of the less dc power dissipation, noise immuned and fast.Thelogic is highly automated. Avoid in large fan outs as it leads to excessive levels oflogic.2.BICMOSIt can be used in high speed applications with large fanout. The economics must bejustified.PSUEDO –NMOSMostly useful in large fan in NOR gates like ROMS,PLA and CLA adders.The DCpower can be reduced to 0 in case of power down situationsClocked CMOSUseful in hot electron susceptible processes.CMOS domino logicUsed mostly in high speed low power application. Care must take of chargeredistribution. Precharge robs the speed advantage.CVSLThis is basically useful in fast cascaded logic .The size, design complexity andreduced noise immunity make the design not so popular.Hybrid designs are also being tried for getting the maximum advantage of eachof them into one.
    • UNIT 4BASIC CIRCUIT DESIGN CONCEPTSINTRODUCTIONWe have already seen that MOS structures are formed by the super imposition of anumber conducting ,insulating and transistor forming material. Now each of theselayers have their own characteristics like capacitance and resistances. Thesefundamental components are required to estimate the performance of the system.These layers also have inductance characteristics that are important for I/O behaviourbut are usually neglected for on chip devices.The issues of prominence are1.Resistance, capacitance and inductance calculations.2.Delay estimations3.Determination of conductor size for power and clock distribution4.Power consumption5.Charge sharing6.Design margin7.Reliabiltiy8.Effects and extent of scalingRESISTANCE ESTIMATIONThe concept of sheet resistance is being used to know the resistive behavior of thelayers that go into formation of the MOS device. Let us consider a uniform slab ofconducting material of the following characteristics .Resistivity- ρWidth - WThickness - tLength between faces – L as shown nextFigure 24:A slab of semiconductor
    • We know that the resistance is given by RAB= ρL/A . The area of the slabconsidered above is given by A=Wt. There fore RAB= ρL/Wt . If the slab isconsidered as a square then L=W. therefore RAB=ρ/t which is called as sheetresistance represented by Rs.The unit of sheet resistance is ohm per square. It is tobe noted that Rs is independent of the area of the slab. Hence we can conclude that a1um per side square has the same resistance as that of 1cm per side square of thesame material.The resistance of the different materials that go into making of the MOS devicedepend on the resistivity and the thickness of the material. For a diffusion layer thedepth defines the thickness and the impurity defines the resistivity. The table of valuesfor a 5u technology is listed below.5u technology means minimum line width is 5uand = 2.5u.The diffusion mentioned in the table is n diffusion, p diffusion values are2.5 times of that of n. The table of standard sheet resistance value follows. Layer Rs per square Metal 0.03 Diffusion n(for 2.5 times 10 to 50 the n ) Silicide 2 to 4 Polysilicon 15 to 100 N transistor gate 104 P transistor gate 2.5x 104SHEET RESISTANCE OF MOS TRANSISTORS Figure 25 Min sized inverter
    • The N transistor above is formed by a 2 wide poly and n diffusion. The L/W ratio is1. Hence the transistor is a square, therefore the resistance R is 1sqxRs ohm/sq i.e.R=1x104. If L/W ratio is 4 then R = 4x104. If it is a P transistor then for L/W =1,thevalue of R is 2.5x104.Pull up to pull down ratio = 4.In this case when the nmos is on, both the devices areon simultaneously, Hence there is an on resistance Ron = 40+10 =50k. It is thisresistance that leads the static power consumption which is the disadvantage of nmosdepletion mode devicesFigure 27: Cmos inverterSince both the devices are not on simultaneously there is no static power dissipation The resistance of non rectangular shapes is a little tedious to estimate. Hence it iseasier to convert the irregular shape into regular rectangular or square blocks and thenestimate the resistance. For exampleFigure 28:Irregular rectangular shapes
    • CONTACT AND VIA RESISTANCEThe contacts and the vias also have resistances that depend on the contacted materialsand the area of contact. As the contact sizes are reduced for scaling ,the associatedresistance increases. The resistances are reduced by making ohmic contacts which arealso called loss less contacts. Currently the values of resistances vary from .25ohms toa few tens of ohms.SILICIDESThe connecting lines that run from one circuit to the other have to be optimized. Forthis reason the width is reduced considerably. With the reduction is width the sheetresistance increases, increasing the RC delay component. With poly silicon the sheetresistance values vary from 15 to 100 ohm. This actually effects the extent of scalingdown process. Polysilicon is being replaced with silicide. Silicide is obtained bydepositing metal on polysilicon and then sintering it. Silicides give a sheet resistanceof 2 to 4 ohm. The reduced sheet resistance makes silicides a very attractivereplacement for poly silicon. But the extra processing steps is an offset to theadvantage.A ProblemA particular layer of MOS circuit has a resistivity ρ of 1 ohm –cm. The section is55um long,5um wide and 1 um thick. Calculate the resistance and also find RsR= RsxL/W, Rs= ρ/tRs=1x10-2/1x10-6=104ohmR= 104x55x10-6/5x106=110kCAPACITANCE ESTIMATIONParasitics capacitances are associated with the MOS device due to different layers thatgo into its formation. Interconnection capacitance can also be formed by the metal,diffusion and polysilicon (these are often called as runners) in addition with thetransistor and conductor resistance. All these capacitances actually define theswitching speed of the MOS device.Understanding the source of parasitics and their variation becomes a very essentialpart of the design specially when system performance is measured in terms of thespeed. The various capacitances that are associated with the CMOS device are1.Gate capacitance - due to other inputs connected to output of the device2.Diffusion capacitance - Drain regions connected to the output3.Routing capacitance- due to connections between output and other inputsThe fabrication process illustrates that the conducting layers are apparently seperatedfrom the substrate and other layers by the insulating layer leading to the formation ofparallel capacitors. Since the silicon dioxide is the insulator knowing its thickness wecan calculate the capacitance C= εoεinsA farad DƐεo= permittivity of free space-8.854x1014f/cm
    • εins= relative permitivity of sio2=4.0D= thickness of the dioxide in cmA = area of the plate in cm2The gate to channel capacitance formed due to the sio2 separation is the mostprofound of the mentioned three types. It is directly connected to the input and theoutput. The other capacitance like the metal, poly can be evaluated against thesubstrate. The gate capacitance is therefore standardized so as to enable to move fromone technology to the other conveniently.The standard unit is denoted by Cg. It represents the capacitance between gate to ロchannel with W=L=min feature size. Here is a figure showing the differentcapacitances that add up to give the total gate capacitanceCgd, Cgs = gate to channel capacitance lumped at the source and drainCsb, Cdb = source and drain diffusion capacitance to substrateCgb = gate to bulk capacitanceTotal gate capacitance Cg = Cgd+Cgs+CgbSince the standard gate capacitance has been defined, the other capacitances likepolysilicon, metal, diffusion can be expressed in terms of the same standard units sothat the total capacitance can be obtained by simply adding all the values. In order toexpress in standard values the following steps must be followed1. Calculate the areas of area under consideration relative to that of standard gatei.e.4 2. (standard gate varies according to the technology)2. Multiply the obtained area by relative capacitance values tabulated .3. This gives the value of the capacitance in the standard unit of capacitance Cg. ロTable 1:Relative value of Cglayer Relative value for5u technologyGate to channel 1Diffusion 0.25Poly to sub 0.1M1 to sub 0.075M2 to sub 0.05M2 to M1 0.1M2 to poly 0.075
    • For a 5u technology the area of the minimum sized transistor is 5uX5u=25um2 ie =2.5u, hence,area of minimum sized transistor in lambda is 2 X 2 = 4 2.Thereforefor 2u or 1.2u or any other technology the area of a minimum sized transistor inlambda is 4 2. Lets solve a few problems to get to know the things better.The figure above shows the dimensions and the interaction of different layers, forevaluating the total capacitance resulting so.Three capacitance to be evaluated metal Cm,polysilicon Cp and gate capacitance CgArea of metal = 100x3=300 2Relative area = 300/4=75Cm=75Xrelative cap=75X0.075=5.625 Cg ロPolysilicon capacitance CpArea of poly=(4x4+1x2+2X2)=22 2Relative area = 22 2/4 2=5.5Cp=5.5Xrelative cap=5.5x.1=0.55 CgロGate capacitance Cg= 1 Cg because it is a min size gate ロCt=Cm+Cp+Cg=5.625+0.55+1=7.2 Cg ロFigure 29:Mos structure
    • The input capacitance is made of three components metal capacitance Cm, polycapacitance Cp, gate capacitance Cg i.e Cin= Cm+Cg+CpRelative area of metal =(50x3)X2/4=300/4=75Cm=75x0.075=5.625 Cg ロRelative area of poly = (4x4+2x1+2x2)/4 =22/4 =5.5Cp=5.5X0.1=0.55 Cg ロCg=1 Cg ロCin=7.175 Cg ロCout = Cd+Cperi. Assuming Cperi to be negligible.Cout = Cd.Relative area of diffusion=51x2/4=102/4=25.5Cd=25.5x0.25=6.25 Cg. ロThe relative values are for the 5um technologyDELAY The concept of sheet resistance and standard unit capacitance can be used tocalculate the delay. If we consider that a one feature size poly is charged by onefeature size diffusion then the delay is Time constant 1Ɛ= Rs (n/p channel)x 1 Cg ロsecs. This can be evaluated for any technology. The value of Cg will vary with ロdifferent technologies because of the variation in the minimum feature size.5u using n diffusion=104X0.01=0.1ns safe delay 0.03nsec2um = 104x0.0032=0.064 nsecs safe delay 0.02nsec1.2u= 104x0.0023 = 0.046nsecs safe delay =0.1nsecThese safe figures are essential in order to anticipate the output at the right timeINVERTER DELAYSWe have seen that the inverter is associated with pull up and pull down resistancevalues. Specially in nmos inverters. Hence the delay associated with the inverter willdepend on whether it is being turned off or on. If we consider two inverters cascadedthen the total delay will remain constant irrespective of the transitions. Nmos andCmos inverter delays are shown nextNMOS INVERTERFigure 30: Cascaded nmos inverters
    • Let us consider the input to be high and hence the first inverter will pull it down. Thepull down inverter is of minimum size nmos. Hence the delay is 1Ɛ. Second inverterwill pull it up and it is 4 times larger, hence its delay is 4Ɛ.The total delay is 1Ɛ+4Ɛ= 5Ɛ. Hence for nmos the delay can be generalized as T=(1+Zpu/Zpd) ƐCMOS INVERTERFigure 30 : Cascaded Cmos inverterLet us consider the input to be high and hence the first inverter will pull it down. Thenmos transistor has Rs = 10k and the capacitance is 2Cg. Hence the delay is 2Ɛ. Nowthe second inverter will pull it up, job done by the pmos. Pmos has sheet resistance of25k i.e 2.5 times more, everything else remains same and hence delay is 5Ɛ. Totaldelay is 2Ɛ +5Ɛ = 7Ɛ . The capacitance here is double because the input isconnected to the common poly, putting both the gate capacitance in parallel. The onlyfactor to be considered is the resistance of the p gate which is increasing the delay. Ifwant to reduce delay, we must reduce resistance. If we increase the width of pchannel, resistance can be reduced but it increases the capacitance. Hence some tradeoff must be made to get the appropriate values.FORMAL ESTIMATION OF DELAYThe inverter either charges or discharges the load capacitance CL. We could alsoestimate the delay by estimating the rise time and fall time theoritically.Rise time estimationAssuming that the p device is in saturation we have the current given by the equationIdsp=ßp(Vgs-|Vtp|)2/2Figure 31 :Rise time estimation
    • The above current charges the capacitance and it has a constant value therefore themodel can be written as shown in figure above. The output is the drop across thecapacitance, given byVout =Idsp x t/CLSubstituting for Idsp we have Vout=ßp(Vgs-|Vtp|)2t/2CL. Therefore the equation fort=2CLVout/ßp(Vgs-|Vtp|).Let t=Ɛr and Vout=Vdd, therefore we have Ɛr =2VddCL/ßp(Vgs-|Vtp|)2. If consider Vtp=0.2Vdd and Vgs=Vdd we have Ɛr =3CL/ßpVddOn similar basis the fall time can be also be written as Ɛf = 3CL/ßnVdd whose modelcan be written as shown nextFigure 32 :Fall time estimationDRIVING LARGE CAPACITIVE LOADThe problem of driving large capacitive loads arises when signals must travel outsidethe chip. Usually it so happens that the capacitance outside the chip are higher. Toreduce the delay these loads must be driven by low resistance. If we are using acascade of inverter as drivers the pull and pull down resistances must be reduced.Low résistance means low L:W ratio. To reduce the ratio, W must be increased. SinceL cannot be reduced to lesser than minimum we end up having a device whichoccupies a larger area. Larger area means the input capacitance increases and slowsdown the process more. The solution to this is to have N cascaded inverters with theirsizes increasing, having the largest to drive the load capacitance. Therefore if we have3 inverters,1st is smallest and third is biggest as shown next.Figure 33:Cascaded inverters with varying widths
    • We see that the width is increasing by a factor of f towards the last stage. Now both fand N can be complementary. If f for each stage is large the number of stages Nreduces but delay per stage increases. Therefore it becomes essential to optimize. FixN and find the minimum value of f. For nmos inverters if the input transitions from 0to 1 the delay is fƐ and if it transitions from 1 to 0 the delay is 4 fƐ. The delay for anmos pair is 5 fƐ. For a cmos pair it will be 7fƐoptimum value of f.Assume y=CL/ Cg = fN, therefore choice of values of N and f are interdependent. ロWe find the value of f to minimize the delay, from the equation of y we haveln(y)=Nln(f) i.e N=ln(y)/ln(f). If delay per stage is 5fƐ for nmos, then for evennumber of stages the total delay is N/2 5fƐ=2.5fƐ. For cmos total delay is N/2 7fƐ =3.5fƐHence delay ά Nft=ln(y)/ln(f)ft. Delay can be minimized if chose the value of f to beequal to e which is the base of natural logarithms. It means that each stage is 2.7widerthan its predecessor. If f=e then N= ln(y).The total delay is then given by1.For N=eventd=2.5NeƐ for nmos, td=3.5NeƐ for cmos2.For N=oddtransition from 0 to 1 transition from1 to 0td=[2.5(N-1)+1]eƐ nmos td=[2.5(N-1)+4]eƐtd=[3.59N-1)+2]eƐ cmos td=[3.5(N-1)+5]eƐfor exampleFor N=5 which is odd we can calculate the delay fro vin=1 as td=[2.5(5-1)+1]eƐ=11e Ɛi.e. 1 +4+1+4+1 = 11eƐFor vin =0 , td=[2.5(5-1)+4]eƐ = 14eƐ4+1+4+1+4 = 14eƐSUPER BUFFERThe asymmetry of the inverters used to solve delay problems is clearly undesirable,this also leads to more delay problems, super buffer are a better solution. We have ainverting and non inverting variants of the super buffer. Such arrangements whenused for 5u technology showed that they were capable of driving 2pf capacitance with2nsec rise time.The figure shown next is the inverting variant.
    • IFigure 34:Inverterting bufferFigure 34:NonInverteing variantBICMOS DRIVERSThe availability of bipolar devices enables us to use these as the output stage ofinverter or any logic. Bipolar devices have high Tran con--ductance and they are ableswitch large currents with smaller input voltage swings. The time required to changethe out by an amount equal to the input is given by ∆t=CL/gm, Where gm is thedevice trans conductance. ∆t will be a very small value because of the high gm. Thetransistor delay consists of two components Tin and TL. Tin the time required tocharge the base of the transistor which is large. TL is smaller because the time taketo charge capacitor is less by hfe which is the transistor gain a comparative graphshown below.
    • Figure 35The collector resistance is another parameter that contributes to the delay.The graphshown below shows that for smaller load capacitance, the delay is manageable but forlarge capacitance, as Rc increases the delay increase drastically.Figure 36By taking certain care during fabrication reasonably good bipolar devices can beproduced with large hfe, gm ,ß and small Rc. Therefore bipolar devices used inbuffers and logic circuits give the designers a lot of scpoe and freedom .This iscoming without having to do any changes with the cmos circuit.PROPAGATION DELAY
    • This is delay introduced when the logic signals have to pass through a chain of passtransistors. The transistors could pose a RC product delay and this increasesdrastically as the number of pass transistor in series increases.As seen from the figurethe response at node V2 is given by CdV2/dt=(V1-V2)(V2-V3)/R For a long networkwe can write RCdv/dt =dv2/dx2, i.e delay ά x2,Figure 38Lump all the R and C we have Rtotal=nrRs and C=nc Cg where and hence delay ロ=n2rcƐ. The increases by the square of the number, hence restrict the number ofstages to maximum 4 and for longer ones introduce buffers in between.DESIGN OF LONG POLYSILICONSThe following points must be considered before going in for long wire.1.The designer is also discouraged from designing long diffusion lines also becausethe capacitance is much larger2.When it inevitable and long poly lines have to used the best way to reduce delay isuse buffers in between. Buffers also reduce the noise sensitivityOTHER SOURCES OF CAPACITANCEWiring capacitance1.Fringing field2.Interlayer capacitance
    • 3.Peripheral capacitanceThe capacitances together add upto as much capacitance as coming from the gate tosource and hence the design must consider points to reduce them.The major of thewiring capacitance is coming from fringing field effects. Fringing capacitances is dueto parallel fine metal lines running across the chip for power conection.Thecapacitance depends on the length l, thickness t and the distance d between the wireand the substrate. The accurate prediction is required for performanceestimation.Hence Cw=Carea+Cff.Interlayer capacitance is seen when different layers cross each and hence it isneglected for simole calculations. Such capacitance can be easily estimated for regularstructures and helps in modeling the circuit better.Peripheral capacitance is seen at the junction of two devices. The source and thedrain n regions form junctions with the pwell (substrate) and p diffusion form withadjacent nwells leading to these side wall (peripheral) capacitanceThe capacitances are profound when the devices are shrunk in sizes and hence mustbe considered. Now the total diffusion capacitance is Ctotal = Carea + CperiIn order to reduce the side wall effects, the designers consider to use isolation regionsof alternate impurity.CHOICE OF LAYERS1.Vdd and Vss lines must be distributed on metal lines except for some exception2.Long lengths of poly must be avoided because they have large Rs,it is not suitablefor routing Vdd or Vss lines.3.Since the resistance effects of the transistors are much larger, hence wiring effectsdue to voltage dividers are not that profoundCapacitance must be accurately calculated for fast signal lines usually those usinghigh Rs material. Diffusion areas must be carefully handled because they have largercapacitance to substrate.With all the above inputs it is better to model wires as small capacitors which willgive electrical guidelines for communication circuits.PROBLEMS1.A particular section of the layout includes a 3 wide metal path which crosses a 2polysilicon path at right angles. Assuming that the layers are seperated by a 0.5 thicksio2,find the capacitance between the two.Capacitance = Ɛ0 Ɛins A/DLet the technology be 5um, =2.5um.Area = 7.5umX5um=37.5umC=4X8.854X10-12 x37.5/ 0.5 =2656pFThe value of C in standard units isRelative area 6 2 /4 2 =1.5C =1.5x0.075=0.1125 Cgロ2 nd part of the problem
    • The polysilicon turns across a 4 diffusion layer, find the gate to channel capacitance.Area = 2 x 4 =8 2 Relative area= 8 2 / 4 2 =2Relative capacitance for 5u=1Total gate capacitance = 2 Cg ロGate to channel capacitance>metal2. The two nmos transistors are cascaded to drive a load capacitance of 16 Cg asロshown in figure ,Calculate the pair delay. What are the ratios of each transistors. fstray and wiring capacitance is to be considered then each inverter will have anadditional capacitance at the output of 4 Cg .Find the delay. ロFigure 40Lpu=16 Wpu=2 Zpu=8Lpd=2 Wpd=2 Zpd=1Ratio of inverter 1 = 8:1Lpu=2 Wpu=2 Zpu=1Lpd =2 Wpd =8 Zpd=1/4Ratio of inverter 2 = 1/1/4=4Delay without strays1Ɛ=Rsx1 CgロLet the input transition from 1 to 0Delay 1 = 8RsX Cg=8Ɛ ロ Delay 2=4Rs( Cg +16 Cg)=68Ɛ Total delay = 76Ɛ ロ ロDelay with straysDelay 1 = 8RsX( Cg+ 4 Cg) = 40Ɛ Delay 2= 4RsX( Cg+ 4 Cg +16 Cg)=84Ɛ ロ ロ ロ ロ ロTotal delay = 40+84=124ƐIf Ɛ = 0.1ns for 5u ie the delays are 7.6ns and 12.4nsSCALING OF MOS DEVICES
    • The VLSI technology is in the process of evolution leading to reduction of the featuresize and line widths. This process is called scaling down. The reduction in sizes hasgenerally lead to better performance of the devices. There are certain limits on scalingand it becomes important to study the effect of scaling. The effect of scaling must bestudied for certain parameters that effect the performance.The parameters are as stated below1.Minimum feature size2.Number of gates on one chip3.Power dissipation4.Maximum operational frequency5.Die size6.Production cost .These are also called as figures of meritMany of the mentioned factors can be improved by shrinking the sizes of transistors,interconnects, separation between devices and also by adjusting the voltage anddoping levels. Therefore it becomes essential for the designers to implement scalingand understand its effects on the performanceThere are three types of scaling models used1.Constant electric field scaling model2.Constant voltage scaling model3.Combined voltage and field modelThe three models make use of two scaling factors 1/ß and 1/ά . 1/ß is chosen as thescaling factor for Vdd, gate oxide thickness D. 1/ ά is chosen as the scaling factor forall the linear dimensions like length, width etc. the figure next shows the dimensionsand their scaling factorsThe following are some simple derivations for scaling down the device parameters1.Gate area AgAg= L x W. Since L & W are scaled down by 1/ ά. Ag is scaled down by 1/ ά22.Gate capacitance per unit areaCo=Ɛo/D, permittivity of sio2 cannot be scaled, hence Co can be scaled 1/1/ß=ß3.Gate capacitance CgCg=CoxA=CoxLxW. Therefore Cg can be scaled by ßx1/ άx1/ ά= ß/ ά24.Parasitic capacitance Cx=Ax/d, where Ax is the area of the depletion around the drain or source. d is thedepletion width .Ax is scaled down by 1/ά2 and d is scaled by 1/ά. Hence Cx isscaled by1/ ά2 /1/ ά = 1/ ά
    • 5.Carrier density in the channel QonQon=Co.VgsCo is scaled by ß and Vgs is scaled by 1/ ß,hence Qo is scaled by ßx1/ß =1.Channel resistance RoRon = L/Wx1/Qoxµ, µ is mobility of charge carriers . Ro is scaled by1/ά/1/ άx1=1Gate delay TdTd is proportional to Ro and CgTd is scaled by 1x ß/ά2 = ß/ά2Maximum operating frequency fofo=1/td,therefore it is scaled by 1/ ß/ά2 = ά2/ßSaturation currentIdss= CoµW(Vgs-Vt)/2L, Co scale by ß andvoltages by 1/ ß, Idss is scaled by ß /ß2= 1/ßCurrent DensityJ=Idss/A hence J is scaled by 1/ß/1/ά2 = ά2 /ß